function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
// Process all other feeds using last_updated and interval parameters
+ $schema_version = get_schema_version($link);
+
+ if ($schema_version != SCHEMA_VERSION) {
+ die("Schema version is wrong, please upgrade the database.\n");
+ }
+
define('PREFS_NO_CACHE', true);
// Test if the user has loggued in recently. If not, it does not update its feeds.
if (DB_TYPE == "pgsql") {
$update_limit_qpart = "AND ((
ttrss_feeds.update_interval = 0
+ AND ttrss_user_prefs.value != '-1'
AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
) OR (
ttrss_feeds.update_interval > 0
} else {
$update_limit_qpart = "AND ((
ttrss_feeds.update_interval = 0
+ AND ttrss_user_prefs.value != '-1'
AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
) OR (
ttrss_feeds.update_interval > 0
$random_qpart = sql_random_function();
// We search for feed needing update.
- $result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id, ttrss_feeds.owner_uid,
- ".SUBSTRING_FOR_DATE."(ttrss_feeds.last_updated,1,19) AS last_updated,
- ttrss_feeds.update_interval
+ $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
FROM
ttrss_feeds, ttrss_users, ttrss_user_prefs
WHERE
AND ttrss_users.id = ttrss_user_prefs.owner_uid
AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
$login_thresh_qpart $update_limit_qpart
- $updstart_thresh_qpart
+ $updstart_thresh_qpart
ORDER BY $random_qpart $query_limit");
$user_prefs_cache = array();
- if($debug) _debug(sprintf("Scheduled %d feeds to update...\n", db_num_rows($result)));
+ if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
// Here is a little cache magic in order to minimize risk of double feed updates.
$feeds_to_update = array();
while ($line = db_fetch_assoc($result)) {
- $feeds_to_update[$line['id']] = $line;
+ array_push($feeds_to_update, db_escape_string($link, $line['feed_url']));
}
// We update the feed last update started date before anything else.
// There is no lag due to feed contents downloads
// It prevent an other process to update the same feed.
- $feed_ids = array_keys($feeds_to_update);
- if($feed_ids) {
+
+ if(count($feeds_to_update) > 0) {
+ $feeds_quoted = array();
+
+ foreach ($feeds_to_update as $feed) {
+ array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'");
+ }
+
db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
- WHERE id IN (%s)", implode(',', $feed_ids)));
+ WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
}
expire_cached_files($debug);
expire_lock_files($debug);
+ expire_error_log($link, $debug);
- // For each feed, we call the feed update function.
- while ($line = array_pop($feeds_to_update)) {
-
- if($debug) _debug("Feed: " . $line["feed_url"] . ", " . $line["last_updated"]);
+ $nf = 0;
- update_rss_feed($link, $line["id"], true);
-
- sleep(1); // prevent flood (FIXME make this an option?)
+ // For each feed, we call the feed update function.
+ foreach ($feeds_to_update as $feed) {
+ if($debug) _debug("Base feed: $feed");
+
+ //update_rss_feed($link, $line["id"], true);
+
+ // since we have the data cached, we can deal with other feeds with the same url
+
+ $tmp_result = db_query($link, "SELECT DISTINCT ttrss_feeds.id,last_updated
+ FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
+ ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
+ ttrss_users.id = ttrss_user_prefs.owner_uid AND
+ ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
+ feed_url = '".db_escape_string($link, $feed)."' AND
+ (ttrss_feeds.update_interval > 0 OR
+ ttrss_user_prefs.value != '-1')
+ $login_thresh_qpart
+ ORDER BY ttrss_feeds.id $query_limit");
+
+ if (db_num_rows($tmp_result) > 0) {
+ while ($tline = db_fetch_assoc($tmp_result)) {
+ if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
+ update_rss_feed($link, $tline["id"], true);
+ ++$nf;
+ }
+ }
}
require_once "digest.php";
// Send feed digests by email if needed.
send_headlines_digests($link, $debug);
+ return $nf;
+
} // function update_daemon_common
// ignore_daemon is not used
$result = db_query($link, "SELECT id,update_interval,auth_login,
feed_url,auth_pass,cache_images,last_updated,
mark_unread_on_update, owner_uid,
- pubsub_state
+ pubsub_state, auth_pass_encrypted
FROM ttrss_feeds WHERE id = '$feed'");
if (db_num_rows($result) == 0) {
$mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
0, "mark_unread_on_update"));
$pubsub_state = db_fetch_result($result, 0, "pubsub_state");
+ $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
+ 0, "auth_pass_encrypted"));
db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
WHERE id = '$feed'");
$auth_login = db_fetch_result($result, 0, "auth_login");
$auth_pass = db_fetch_result($result, 0, "auth_pass");
+ if ($auth_pass_encrypted) {
+ require_once "crypt.php";
+ $auth_pass = decrypt_string($auth_pass);
+ }
+
$cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
$fetch_url = db_fetch_result($result, 0, "feed_url");
$feed = db_escape_string($link, $feed);
- /* if ($auth_login && $auth_pass ){
- $url_parts = array();
- preg_match("/(^[^:]*):\/\/(.*)/", $fetch_url, $url_parts);
+ if ($override_url) $fetch_url = $override_url;
- if ($url_parts[1] && $url_parts[2]) {
- $fetch_url = $url_parts[1] . "://$auth_login:$auth_pass@" . $url_parts[2];
- }
- } */
+ $date_feed_processed = date('Y-m-d H:i');
- if ($override_url)
- $fetch_url = $override_url;
+ $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
+
+ // Ignore cache if new feed or manual update.
+ $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
+ 30 : get_feed_update_interval($link, $feed) * 60;
if ($debug_enabled) {
- _debug("update_rss_feed: fetching [$fetch_url]...");
+ _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
+ _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
}
- // Ignore cache if new feed or manual update.
- $cache_age = (is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
- -1 : get_feed_update_interval($link, $feed) * 60;
+ $cached_feed_data_hash = false;
+
+ $rss = false;
+ $rss_hash = false;
+ $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
+ $last_updated_timestamp = strtotime($last_updated);
- $simplepie_cache_dir = CACHE_DIR . "/simplepie";
+ $force_refetch = isset($_REQUEST["force_refetch"]);
- if (!is_dir($simplepie_cache_dir)) {
- mkdir($simplepie_cache_dir);
+ if (file_exists($cache_filename) &&
+ is_readable($cache_filename) &&
+ !$auth_login && !$auth_pass &&
+ filemtime($cache_filename) > time() - $cache_age) {
+
+ if ($debug_enabled) {
+ _debug("update_rss_feed: using local cache.");
+ }
+
+ if ($cache_timestamp > $last_updated_timestamp) {
+ @$rss_data = file_get_contents($cache_filename);
+
+ if ($rss_data) {
+ $rss_hash = sha1($rss_data);
+ @$rss = unserialize($rss_data);
+ }
+ } else if (!$force_refetch) {
+ if ($debug_enabled) {
+ _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
+ }
+ return;
+ }
}
- $feed_data = fetch_file_contents($fetch_url, false,
- $auth_login, $auth_pass, false, $no_cache ? 15 : 45);
+ if (!$rss) {
- if (!$feed_data) {
- global $fetch_last_error;
+ if (!$feed_data) {
+ if ($debug_enabled) {
+ _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
+ }
+
+ $feed_data = fetch_file_contents($fetch_url, false,
+ $auth_login, $auth_pass, false,
+ $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
+ $force_refetch ? 0 : max($last_updated_timestamp, $cache_timestamp));
+
+ if ($debug_enabled) {
+ _debug("update_rss_feed: fetch done.");
+ }
- if ($debug_enabled) {
- _debug("update_rss_feed: unable to fetch: $fetch_last_error");
}
- $error_escaped = db_escape_string($link, $fetch_last_error);
+ if (!$feed_data) {
+ global $fetch_last_error;
+ global $fetch_last_error_code;
- db_query($link,
- "UPDATE ttrss_feeds SET last_error = '$error_escaped',
- last_updated = NOW() WHERE id = '$feed'");
+ if ($debug_enabled) {
+ _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
+ }
+
+ $error_escaped = '';
+
+ // If-Modified-Since
+ if ($fetch_last_error_code != 304) {
+ $error_escaped = db_escape_string($link, $fetch_last_error);
+ } else {
+ if ($debug_enabled) {
+ _debug("update_rss_feed: source claims data not modified, nothing to do.");
+ }
+ }
+
+ db_query($link,
+ "UPDATE ttrss_feeds SET last_error = '$error_escaped',
+ last_updated = NOW() WHERE id = '$feed'");
- return;
+ return;
+ }
}
$pluginhost = new PluginHost($link);
$feed_data = $plugin->hook_feed_fetched($feed_data);
}
- if ($debug_enabled) {
- _debug("update_rss_feed: fetch done, parsing...");
- }
-
- $rss = new SimplePie();
- $rss->set_sanitize_class("SanitizeDummy");
- // simplepie ignores the above and creates default sanitizer anyway,
- // so let's override it...
- $rss->sanitize = new SanitizeDummy();
- $rss->set_output_encoding('UTF-8');
- $rss->set_raw_data($feed_data);
-
- if ($debug_enabled) {
- _debug("feed update interval (sec): " .
- get_feed_update_interval($link, $feed)*60);
- }
-
- $rss->enable_cache(!$no_cache);
-
- if (!$no_cache) {
- $rss->set_cache_location($simplepie_cache_dir);
- $rss->set_cache_duration($cache_age);
+ if (!$rss) {
+ $rss = new SimplePie();
+ $rss->set_sanitize_class("SanitizeDummy");
+ // simplepie ignores the above and creates default sanitizer anyway,
+ // so let's override it...
+ $rss->sanitize = new SanitizeDummy();
+ $rss->set_output_encoding('UTF-8');
+ $rss->set_raw_data($feed_data);
+ $rss->enable_cache(false);
+
+ @$rss->init();
}
- @$rss->init();
-
// print_r($rss);
$feed = db_escape_string($link, $feed);
if (!$rss->error()) {
+ // cache data for later
+ if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
+ $rss_data = serialize($rss);
+ $new_rss_hash = sha1($rss_data);
+
+ if ($new_rss_hash != $rss_hash) {
+ if ($debug_enabled) {
+ _debug("update_rss_feed: saving $cache_filename");
+ }
+ @file_put_contents($cache_filename, serialize($rss));
+ }
+ }
+
// We use local pluginhost here because we need to load different per-user feed plugins
$pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
$site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
- if ($debug_enabled) {
- _debug("update_rss_feed: checking favicon...");
- }
+ if ($favicon_needs_check || $force_refetch) {
+ if ($debug_enabled) {
+ _debug("update_rss_feed: checking favicon...");
+ }
- if ($favicon_needs_check) {
check_feed_favicon($site_url, $feed, $link);
+ $favicon_file = ICONS_DIR . "/$feed.ico";
+
+ if (file_exists($favicon_file)) {
+ require_once "colors.php";
+
+ $favicon_color = db_escape_string($link,
+ calculate_avg_color($favicon_file));
+
+ $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
+ }
db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
+ $favicon_colorstring
WHERE id = '$feed'");
}
if (!$entry_guid) $entry_guid = $item->get_link();
if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
- if ($debug_enabled) {
- _debug("update_rss_feed: guid $entry_guid");
- }
-
if (!$entry_guid) continue;
$entry_guid = "$owner_uid,$entry_guid";
+ $entry_guid_hashed = db_escape_string($link, 'SHA1:' . sha1($entry_guid));
+
+ if ($debug_enabled) {
+ _debug("update_rss_feed: guid $entry_guid / $entry_guid_hashed");
+ }
+
$entry_timestamp = "";
$entry_timestamp = strtotime($item->get_date());
- if ($entry_timestamp == -1 || !$entry_timestamp) {
+ if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
$entry_timestamp = time();
$no_orig_date = 'true';
} else {
_debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
}
- $entry_title = $item->get_title();
+ $entry_title = html_entity_decode($item->get_title());
$entry_link = rewrite_relative_url($site_url, $item->get_link());
// FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
$result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
- WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
+ WHERE ref_id = id AND (guid = '".db_escape_string($link, $entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
if (db_num_rows($result) != 0) {
$entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
db_query($link, "BEGIN");
$result = db_query($link, "SELECT id FROM ttrss_entries
- WHERE guid = '$entry_guid'");
+ WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
if (db_num_rows($result) == 0) {
author)
VALUES
('$entry_title',
- '$entry_guid',
+ '$entry_guid_hashed',
'$entry_link',
'$entry_timestamp_fmt',
'$entry_content',
'',
$no_orig_date,
NOW(),
- NOW(),
+ '$date_feed_processed',
'$entry_comments',
'$num_comments',
'$entry_plugin_data',
// now it should exist, if not - bad luck then
$result = db_query($link, "SELECT
- id,content_hash,no_orig_date,title,plugin_data,
+ id,content_hash,no_orig_date,title,plugin_data,guid,
".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
num_comments
FROM
ttrss_entries
- WHERE guid = '$entry_guid'");
+ WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
$entry_ref_id = 0;
$entry_int_id = 0;
if (db_num_rows($result) == 1) {
if ($debug_enabled) {
- _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
+ _debug("update_rss_feed: base guid found, checking for user record");
}
// this will be used below in update handler
$ref_id = db_fetch_result($result, 0, "id");
$entry_ref_id = $ref_id;
+ /* $stored_guid = db_fetch_result($result, 0, "guid");
+ if ($stored_guid != $entry_guid_hashed) {
+ if ($debug_enabled) _debug("upgrading compat guid to hashed one");
+
+ db_query($link, "UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
+ id = '$ref_id'");
+ } */
+
// check for user post link to main table
// do we allow duplicate posts with same GUID in different feeds?
if ($post_needs_update) {
if (defined('DAEMON_EXTENDED_DEBUG')) {
- _debug("update_rss_feed: post $entry_guid needs update...");
+ _debug("update_rss_feed: post $entry_guid_hashed needs update...");
}
// print "<!-- post $orig_title needs update : $post_needs_update -->";
return $doc->saveXML($node);
}
+ function expire_error_log($link, $debug) {
+ if ($debug) _debug("Removing old error log entries...");
+
+ if (DB_TYPE == "pgsql") {
+ db_query($link, "DELETE FROM ttrss_error_log
+ WHERE created_at < NOW() - INTERVAL '7 days'");
+ } else {
+ db_query($link, "DELETE FROM ttrss_error_log
+ WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
+ }
+
+ }
+
function expire_lock_files($debug) {
if ($debug) _debug("Removing old lock files...");
}
function expire_cached_files($debug) {
- foreach (array("simplepie", "images", "export") as $dir) {
+ foreach (array("simplepie", "images", "export", "upload") as $dir) {
$cache_dir = CACHE_DIR . "/$dir";
if ($debug) _debug("Expiring $cache_dir");
if ($filter_match) {
foreach ($filter["actions"] AS $action) {
array_push($matches, $action);
+
+ // if Stop action encountered, perform no further processing
+ if ($action["type"] == "stop") return $matches;
}
}
}